Symmetries and continuous attractors in disordered neural circuits.

Disordered neural circuits with circular symmetry can function as continuous attractors, as shown by data-driven recurrent networks modeling rodent head-direction cells.

• Why it matters: Understanding how heterogeneous neural representations relate to idealized models is crucial for deciphering brain function, especially since real neural circuits often deviate from symmetry.
• What they did: The authors constructed recurrent networks from head-direction cell data, reproducing heterogeneous tuning while revealing underlying circular symmetry and continuous-attractor features through large N modeling and mean-field theory.
• The result: Findings suggest that complex mammalian circuits may utilize continuous-attractor mechanisms, with disordered weights carrying symmetry that supports stable bump states, advancing insights into neural representation organization.